Peter Dayan’s research while affiliated with University of Tübingen and other places

Why do today, what you can fail to do tomorrow? Pacing styles involving postponing, and ultimately procrastinating tasks are widespread pathologies to which many succumb. Previous research has recognised multiple types of delayed working, influenced by a myriad of psychological and situational factors. However, the main mechanistic explanation for delays, which concerns temporal discounting of distant rewards, only encompasses a subset. Further, investigations of pacing have been rather independent from those of procrastination. Here, we introduce a systematic taxonomy of pacing and procrastination within a common framework that is based on the type of temporal decisions involved in choosing to delay. We suggest that these decisions are driven by characteristics of the task and the (sub-)optimality of the decision-making. We illustrate aspects of the taxonomy by simulating diverse sources of behavioral delay using reinforcement learning models. We also analyse whether students pacing their work through a semester in a real-world task (Zhang & Ma, Scientific Reports, 2024) showany evidence of behaving according to the types detailed in our taxonomy. Our approach provides a theoretical foundation for understanding pacing and procrastination, enabling the integration of both established and novel mechanisms within a unified conceptual framework.

Animals integrate knowledge about how the state of the environment evolves to choose actions that maximise reward. Such goal-directed behaviour - or model-based (MB) reinforcement learning (RL) - can flexibly adapt choice to changes, being thus distinct from simpler habitual - or model-free (MF) RL - strategies. Previous inactivation and neuroimaging work implicates prefrontal cortex (PFC) and the caudate striatal region in MB-RL; however, details are scarce about its implementation at the single-neuron level. Here, we recorded from two PFC regions - the dorsal anterior cingulate cortex (ACC) and dorsolateral PFC (DLPFC), and two striatal regions, caudate and putamen - while two rhesus macaques performed a sequential decision-making (two-step) task in which MB-RL involves knowledge about the statistics of reward and state transitions. All four regions, but particularly the ACC, encoded the rewards received and tracked the probabilistic state transitions that occurred. However, ACC (and to a lesser extent caudate) encoded the key variables of the task - namely the interaction between reward, transition and choice - which underlies MB decision-making. ACC and caudate neurons also encoded MB-derived estimates of choice values. Moreover, caudate value estimates of the choice options flipped when a rare transition occurred, demonstrating value update based on structural knowledge of the task. The striatal regions were unique (relative to PFC) in encoding the current and previous rewards with opposing polarities, reminiscent of dopaminergic neurons, and indicative of a MF prediction error. Our findings provide a deeper understanding of selective and temporally dissociable neural mechanisms underlying goal-directed behaviour.

Navigating uncertainty is crucial for survival, with the location and availability of reward varying in different and unsignalled ways. Hippocampal place cell populations over-represent salient locations in an animal's environment, including those associated with rewards; however, how the spatial uncertainties impact the cognitive map is unclear. We report a virtual spatial navigation task designed to test the impact of different levels and types of uncertainty about reward on place cell populations. When the reward location changed on a trial-by-trial basis, inducing expected uncertainty, a greater proportion of place cells followed along, and the reward and the track end became anchors of a warped spatial metric. When the reward location then unexpectedly moved, the fraction of reward place cells that followed was greater when starting from a state of expected, compared to low, uncertainty. Overall, we show that different forms of potentially interacting uncertainty generate remapping in parallel, task-relevant, reference frames.

Humans excel at learning abstract patterns across different sequences, filtering out irrelevant details, and transferring these generalized concepts to new sequences. In contrast, many sequence learning models lack the ability to abstract, which leads to memory inefficiency and poor transfer. We introduce a non-parametric hierarchical variable learning model (HVM) that learns chunks from sequences and abstracts contextually similar chunks as variables. HVM efficiently organizes memory while uncovering abstractions, leading to compact sequence representations. When learning on language datasets such as babyLM, HVM learns a more efficient dictionary than standard compression algorithms such as Lempel-Ziv. In a sequence recall task requiring the acquisition and transfer of variables embedded in sequences, we demonstrate HVM's sequence likelihood correlates with human recall times. In contrast, large language models (LLMs) struggle to transfer abstract variables as effectively as humans. From HVM's adjustable layer of abstraction, we demonstrate that the model realizes a precise trade-off between compression and generalization. Our work offers a cognitive model that captures the learning and transfer of abstract representations in human cognition and differentiates itself from the behavior of large language models.

Establishing a unified theory of cognition has been a major goal of psychology. While there have been previous attempts to instantiate such theories by building computational models, we currently do not have one model that captures the human mind in its entirety. Here we introduce Centaur, a computational model that can predict and simulate human behavior in any experiment expressible in natural language. We derived Centaur by finetuning a state-of-the-art language model on a novel, large-scale data set called Psych-101. Psych-101 reaches an unprecedented scale, covering trial-by-trial data from over 60,000 participants performing over 10,000,000 choices in 160 experiments. Centaur not only captures the behavior of held-out participants better than existing cognitive models, but also generalizes to new cover stories, structural task modifications, and entirely new domains. Furthermore, we find that the model's internal representations become more aligned with human neural activity after finetuning. Taken together, Centaur is the first real candidate for a unified model of human cognition. We anticipate that it will have a disruptive impact on the cognitive sciences, challenging the existing paradigm for developing computational models.

Compositional representations are thought to enable humans to generalize across combinatorially vast state spaces. Models with learnable object slots, which encode information about objects in separate latent codes, have shown promise for this type of generalization but rely on strong architectural priors. Models with distributed representations, on the other hand, use overlapping, potentially entangled neural codes, and their ability to support compositional generalization remains underexplored. In this paper we examine whether distributed models can develop linearly separable representations of objects, like slotted models, through unsupervised training on videos of object interactions. We show that, surprisingly, models with distributed representations often match or outperform models with object slots in downstream prediction tasks. Furthermore, we find that linearly separable object representations can emerge without object-centric priors, with auxiliary objectives like next-state prediction playing a key role. Finally, we observe that distributed models' object representations are never fully disentangled, even if they are linearly separable: Multiple objects can be encoded through partially overlapping neural populations while still being highly separable with a linear classifier. We hypothesize that maintaining partially shared codes enables distributed models to better compress object dynamics, potentially enhancing generalization.

Humans need to be on their toes when interacting with competitive others to avoid being taken advantage of. Too much caution out of context can, however, be detrimental and produce false beliefs of intended harm. Here, we offer a formal account of this phenomenon through the lens of Theory of Mind. We simulate agents of different depths of mentalizing within a simple game theoretic paradigm and show how, if aligned well, deep recursive mentalization gives rise to both successful deception as well as reasonable skepticism. However, we also show that if a self is mentalizing too deeply – hyper-mentalizing – false beliefs arise that a partner is trying to trick them maliciously, resulting in a material loss to the self. Importantly, we show that this is only true when hypermentalizing agents believe observed actions are generated intentionally. This theory offers a potential cognitive mechanism for suspiciousness, paranoia, and conspiratorial ideation. Rather than a deficit in Theory of Mind, paranoia may arise from the application of overly strategic thinking to ingenuous behaviour. Author Summary Interacting competitively requires vigilance to avoid deception. However, excessive caution can have adverse effects, stemming from false beliefs of intentional harm. So far there is no formal cognitive account of what may cause this suspiciousness. Here we present an examination of this phenomenon through the lens of Theory of Mind – the cognitive ability to consider the beliefs, intentions, and desires of others. By simulating interacting computer agents we illustrate how well-aligned agents can give rise to successful deception and justified skepticism. Crucially, we also reveal that overly cautious agents develop false beliefs that an ingenuous partner is attempting malicious trickery, leading to tangible losses. As well as formally defining a plausible mechanism for suspiciousness, paranoia, and conspiratorial thinking, our theory indicates that rather than a deficit in Theory of Mind, paranoia may involve an over-application of strategy to genuine behaviour.

Generalising information from ourselves to others, and others to ourselves allows for both a dependable source of navigation and adaptability in interpersonal exchange. Disturbances to social development in sensitive periods can cause enduring and distressing damage to lasting healthy relationships. However, identifying the mechanisms of healthy exchange has been difficult. We introduce a theory of self-other generalisation tested with data from a three-phase social value orientation task - the Intentions Game. We involved individuals with (n=50) and without (n=53) a diagnosis of borderline personality disorder and assessed whether self-other information generalisation may explain interpersonal (in)stability. Healthy controls initially used their preferences to predict others and were influenced by their partners, leading to self-other convergence. In contrast, individuals with borderline personality disorder maintained distinct self-other representations, generating a new neutral prior to begin learning. Both groups steadily reduced their updating over time, with healthy participants showing increased sensitivity to update beliefs. Furthermore, we explored theory-driven individual differences underpinning learning. Overall, the findings provide a clear explanation of how self-other generalisation constrains and assists learning, how childhood adversity disrupts this through separation of internalised beliefs and makes clear predictions about the mechanisms of social information integration under uncertainty.

Failing to make decisions that would actively avoid negative outcomes is central to helplessness. In a Bayesian framework, deciding whether to act is informed by beliefs about the world that can be characterised as priors. However, these priors have not been previously quantified. Here we administered two tasks in which participants decided whether to attempt active avoidance actions. The tasks differed in framing and valence, allowing us to test whether the prior generating biases in behaviour is problem-specific or task-independent and general. We performed extensive comparisons of models offering different structural explanations of the data, finding that a Bayesian model with a task-invariant prior for active avoidance provided the best fit to participants’ trial-by-trial behaviour. The parameters of this prior were reliable, and participants with an optimistic prior also reported higher levels of positive affect. These results show that individual differences in prior beliefs can explain decisions to engage in active avoidance of negative outcomes, providing evidence for a Bayesian conceptualization of helplessness.